GRN1 Cancellation Announcement

[photon]

@Taek

When the phase-out of Cuckatoo31 was announced, Obelisk had to decide whether to abandon the current design and switch to building a Cuckatoo32 capable miner

The original idea was to start with Cuckatoo32 and keep going up to make sure single-die chip would not be made.

Meeting Notes: Governance, Sep 11 2018

Meeting notes Governance meeting held Sep 11 @ 3PM UTC in grin/Lobby on Gitter, full chat transcript here. Grin Council attendance: antioch hashmap igno.peverell jaspervdm lehnberg tromp quentinlesceller yeastplume Agenda points & Actions 1. Action point follow ups from previous meetings 1.1 Mu-sig wallet Not yet set up Suggestion to use Electrum. [ ] Action on @igno.peverell to investigate and set up if suitable. 1.2 Risk mgmt document No update. [ ] Action on @lehnberg to work wi…

so the default option was C32 from Sept 2018

Scheduled PoW upgrades proposal

A major motivation for memory hard PoWs is to limit performance by the memory IO bottleneck. If the memory requirements are set too low, as for instance in the case of Litecoin’s scrypt, then the resulting single-chip ASICs are not meaningfully different from pure computational ones like Bitcoin’s sha256. The same destiny befell Zcash’ Equihash, although they missed the opportunity to set much larger memory requirements at launch. Only ethash has so far succesfully resisted single chip ASICs. S…

Then it was dropped to C31 Nov’18 so it must have been at this point when Obelisk committed to gamble with single-die C31 design so not sure about your timeline there.

Speculation

In my opinion, Obelisk planned single-die chip ASIC before the anti-single-chip (or ASIC in general) measures were taken. David Vorick tried to push the design back to single-chip realm many times on gitter as this was likely the only option for a smaller company. And when C32+ was downgraded to C31+, there was a chance the design could still be used. This is where Obelisk had to either drop the Cuckoo SRAM design and start from zero or risk it with a massive monolithic die and push their way to C32 die.

So Obelisk possibly gambled twice: first when developing ASIC for Cuckoo29 that would have creamed all CPU/GPU miners with single-chip considered impossible by Grin devs at the time. And then when they saw a spark of hope with C32 → C31 change. Not sure I like the implied negativity towards Grin PoW design from Obelisk.